diff --git a/backend/.env.example b/.env.example
similarity index 100%
rename from backend/.env.example
rename to .env.example
diff --git a/backend/.gitignore b/.gitignore
similarity index 100%
rename from backend/.gitignore
rename to .gitignore
diff --git a/backend/Dockerfile b/Dockerfile
similarity index 100%
rename from backend/Dockerfile
rename to Dockerfile
diff --git a/README.md b/README.md
index 0a0e6384afa09bb59da2773ca7a8836846639233..c34b8d529d8807b1ae8af712a7d480bdd1525fd1 100644
--- a/README.md
+++ b/README.md
@@ -1,13 +1,63 @@
----
-title: Matchlens API
-emoji: 📉
-colorFrom: blue
-colorTo: yellow
-sdk: gradio
-sdk_version: 6.31.0
-python_version: '3.12'
-app_file: app.py
-pinned: false
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# MatchLens backend
+
+This FastAPI service powers the existing React replay UI. The initial data mode is
+explicitly a **historical replay**, not a live feed. Cricsheet JSON files can be
+placed in `app/data/matches`; the normalized parser loads them at startup.
+
+## Load multiple real matches
+
+The backend does not scrape an HTML scoreboard. It uses Cricsheet's structured,
+historical ball-by-ball JSON source, which is more reliable for replay and
+statistics. From the `backend` directory, download a bounded set of recent
+men's T20 matches:
+
+```powershell
+python scripts\download_cricsheet.py --count 12
+```
+
+Restart Uvicorn after downloading. `GET /api/matches` will then expose the
+downloaded matches alongside the bundled fallback. The downloader is
+repeatable and only stores selected JSON files locally; do not commit the
+downloaded match data unless your distribution policy permits it.
+
+## Run locally
+
+```powershell
+cd backend
+python -m venv .venv
+.venv\Scripts\Activate.ps1
+pip install -r requirements.txt
+Copy-Item .env.example .env
+uvicorn app.main:app --reload
+```
+
+OpenAPI is available at `http://localhost:8000/docs`. Configure `GEMINI_API_KEY`
+for grounded natural-language explanations. Without it, the API returns a
+deterministic evidence-based fallback and explicitly warns that Gemini is not
+configured. Configure Qdrant before running the ingestion tooling; rule
+documents are never used to store changing match scores.
+
+## Frontend
+
+Use the real API by setting:
+
+```env
+VITE_DEMO_MODE=false
+VITE_API_BASE_URL=http://localhost:8000
+```
+
+The frontend then calls `/api/matches`, `/api/matches/{id}/state`,
+`/api/matches/{id}/timeline`, replay, catch-up, and assistant routes. Demo
+fixtures remain available when `VITE_DEMO_MODE` is not `false`.
+
+## Offline experiments and deployment
+
+`app/fine_tuning/train.py` contains the PEFT/LoRA training entry point and
+`evaluate.py` reports measured accuracy, macro-F1, and a confusion matrix.
+Training is intentionally offline and the API does not train at startup.
+`app/rag/ingestion.py` prepares idempotent document chunks for Qdrant.
+
+Build the Cloud Run image with `docker build -t matchlens-api backend` and pass
+secrets as environment variables. Netlify should use the same
+`VITE_API_BASE_URL` setting and an SPA fallback for `index.html`. Render can
+run the same Docker image with the configured `PORT`.
diff --git a/backend/app.py b/app.py
similarity index 100%
rename from backend/app.py
rename to app.py
diff --git a/backend/app/__init__.py b/app/__init__.py
similarity index 100%
rename from backend/app/__init__.py
rename to app/__init__.py
diff --git a/backend/app/ai/__init__.py b/app/ai/__init__.py
similarity index 100%
rename from backend/app/ai/__init__.py
rename to app/ai/__init__.py
diff --git a/backend/app/ai/graph.py b/app/ai/graph.py
similarity index 100%
rename from backend/app/ai/graph.py
rename to app/ai/graph.py
diff --git a/backend/app/api.py b/app/api.py
similarity index 100%
rename from backend/app/api.py
rename to app/api.py
diff --git a/backend/app/config.py b/app/config.py
similarity index 100%
rename from backend/app/config.py
rename to app/config.py
diff --git a/backend/app/data/__init__.py b/app/data/__init__.py
similarity index 100%
rename from backend/app/data/__init__.py
rename to app/data/__init__.py
diff --git a/backend/app/data/knowledge/cricket_rules_and_laws.txt b/app/data/knowledge/cricket_rules_and_laws.txt
similarity index 100%
rename from backend/app/data/knowledge/cricket_rules_and_laws.txt
rename to app/data/knowledge/cricket_rules_and_laws.txt
diff --git a/backend/app/data/knowledge/t20_analytics_and_chasing_dynamics.txt b/app/data/knowledge/t20_analytics_and_chasing_dynamics.txt
similarity index 100%
rename from backend/app/data/knowledge/t20_analytics_and_chasing_dynamics.txt
rename to app/data/knowledge/t20_analytics_and_chasing_dynamics.txt
diff --git a/backend/app/data/matches/.gitkeep b/app/data/matches/.gitkeep
similarity index 100%
rename from backend/app/data/matches/.gitkeep
rename to app/data/matches/.gitkeep
diff --git a/backend/app/data/matches/1082591.json b/app/data/matches/1082591.json
similarity index 100%
rename from backend/app/data/matches/1082591.json
rename to app/data/matches/1082591.json
diff --git a/backend/app/data/matches/1082592.json b/app/data/matches/1082592.json
similarity index 100%
rename from backend/app/data/matches/1082592.json
rename to app/data/matches/1082592.json
diff --git a/backend/app/data/matches/1082593.json b/app/data/matches/1082593.json
similarity index 100%
rename from backend/app/data/matches/1082593.json
rename to app/data/matches/1082593.json
diff --git a/backend/app/data/matches/1082594.json b/app/data/matches/1082594.json
similarity index 100%
rename from backend/app/data/matches/1082594.json
rename to app/data/matches/1082594.json
diff --git a/backend/app/data/matches/1082595.json b/app/data/matches/1082595.json
similarity index 100%
rename from backend/app/data/matches/1082595.json
rename to app/data/matches/1082595.json
diff --git a/backend/app/data/matches/1082596.json b/app/data/matches/1082596.json
similarity index 100%
rename from backend/app/data/matches/1082596.json
rename to app/data/matches/1082596.json
diff --git a/backend/app/data/matches/1082597.json b/app/data/matches/1082597.json
similarity index 100%
rename from backend/app/data/matches/1082597.json
rename to app/data/matches/1082597.json
diff --git a/backend/app/data/matches/1082598.json b/app/data/matches/1082598.json
similarity index 100%
rename from backend/app/data/matches/1082598.json
rename to app/data/matches/1082598.json
diff --git a/backend/app/data/matches/1082599.json b/app/data/matches/1082599.json
similarity index 100%
rename from backend/app/data/matches/1082599.json
rename to app/data/matches/1082599.json
diff --git a/backend/app/data/matches/1082600.json b/app/data/matches/1082600.json
similarity index 100%
rename from backend/app/data/matches/1082600.json
rename to app/data/matches/1082600.json
diff --git a/backend/app/data/matches/1082601.json b/app/data/matches/1082601.json
similarity index 100%
rename from backend/app/data/matches/1082601.json
rename to app/data/matches/1082601.json
diff --git a/backend/app/data/matches/1082602.json b/app/data/matches/1082602.json
similarity index 100%
rename from backend/app/data/matches/1082602.json
rename to app/data/matches/1082602.json
diff --git a/backend/app/data/matches/__init__.py b/app/data/matches/__init__.py
similarity index 100%
rename from backend/app/data/matches/__init__.py
rename to app/data/matches/__init__.py
diff --git a/backend/app/fine_tuning/__init__.py b/app/fine_tuning/__init__.py
similarity index 100%
rename from backend/app/fine_tuning/__init__.py
rename to app/fine_tuning/__init__.py
diff --git a/backend/app/fine_tuning/data/intent_examples.jsonl b/app/fine_tuning/data/intent_examples.jsonl
similarity index 100%
rename from backend/app/fine_tuning/data/intent_examples.jsonl
rename to app/fine_tuning/data/intent_examples.jsonl
diff --git a/backend/app/fine_tuning/evaluate.py b/app/fine_tuning/evaluate.py
similarity index 100%
rename from backend/app/fine_tuning/evaluate.py
rename to app/fine_tuning/evaluate.py
diff --git a/backend/app/fine_tuning/train.py b/app/fine_tuning/train.py
similarity index 100%
rename from backend/app/fine_tuning/train.py
rename to app/fine_tuning/train.py
diff --git a/backend/app/main.py b/app/main.py
similarity index 100%
rename from backend/app/main.py
rename to app/main.py
diff --git a/backend/app/models.py b/app/models.py
similarity index 100%
rename from backend/app/models.py
rename to app/models.py
diff --git a/backend/app/rag/__init__.py b/app/rag/__init__.py
similarity index 100%
rename from backend/app/rag/__init__.py
rename to app/rag/__init__.py
diff --git a/backend/app/rag/ingestion.py b/app/rag/ingestion.py
similarity index 100%
rename from backend/app/rag/ingestion.py
rename to app/rag/ingestion.py
diff --git a/backend/app/services/__init__.py b/app/services/__init__.py
similarity index 100%
rename from backend/app/services/__init__.py
rename to app/services/__init__.py
diff --git a/backend/app/services/cricket.py b/app/services/cricket.py
similarity index 100%
rename from backend/app/services/cricket.py
rename to app/services/cricket.py
diff --git a/backend/app/services/gemini.py b/app/services/gemini.py
similarity index 100%
rename from backend/app/services/gemini.py
rename to app/services/gemini.py
diff --git a/backend/app/services/intent.py b/app/services/intent.py
similarity index 100%
rename from backend/app/services/intent.py
rename to app/services/intent.py
diff --git a/backend/app/services/match_data.py b/app/services/match_data.py
similarity index 100%
rename from backend/app/services/match_data.py
rename to app/services/match_data.py
diff --git a/backend/app/services/rag.py b/app/services/rag.py
similarity index 100%
rename from backend/app/services/rag.py
rename to app/services/rag.py
diff --git a/backend/app/services/sample_data.py b/app/services/sample_data.py
similarity index 100%
rename from backend/app/services/sample_data.py
rename to app/services/sample_data.py
diff --git a/backend/app/services/stories.py b/app/services/stories.py
similarity index 100%
rename from backend/app/services/stories.py
rename to app/services/stories.py
diff --git a/backend/app/services/workflow.py b/app/services/workflow.py
similarity index 100%
rename from backend/app/services/workflow.py
rename to app/services/workflow.py
diff --git a/backend/README.md b/backend/README.md
deleted file mode 100644
index c34b8d529d8807b1ae8af712a7d480bdd1525fd1..0000000000000000000000000000000000000000
--- a/backend/README.md
+++ /dev/null
@@ -1,63 +0,0 @@
-# MatchLens backend
-
-This FastAPI service powers the existing React replay UI. The initial data mode is
-explicitly a **historical replay**, not a live feed. Cricsheet JSON files can be
-placed in `app/data/matches`; the normalized parser loads them at startup.
-
-## Load multiple real matches
-
-The backend does not scrape an HTML scoreboard. It uses Cricsheet's structured,
-historical ball-by-ball JSON source, which is more reliable for replay and
-statistics. From the `backend` directory, download a bounded set of recent
-men's T20 matches:
-
-```powershell
-python scripts\download_cricsheet.py --count 12
-```
-
-Restart Uvicorn after downloading. `GET /api/matches` will then expose the
-downloaded matches alongside the bundled fallback. The downloader is
-repeatable and only stores selected JSON files locally; do not commit the
-downloaded match data unless your distribution policy permits it.
-
-## Run locally
-
-```powershell
-cd backend
-python -m venv .venv
-.venv\Scripts\Activate.ps1
-pip install -r requirements.txt
-Copy-Item .env.example .env
-uvicorn app.main:app --reload
-```
-
-OpenAPI is available at `http://localhost:8000/docs`. Configure `GEMINI_API_KEY`
-for grounded natural-language explanations. Without it, the API returns a
-deterministic evidence-based fallback and explicitly warns that Gemini is not
-configured. Configure Qdrant before running the ingestion tooling; rule
-documents are never used to store changing match scores.
-
-## Frontend
-
-Use the real API by setting:
-
-```env
-VITE_DEMO_MODE=false
-VITE_API_BASE_URL=http://localhost:8000
-```
-
-The frontend then calls `/api/matches`, `/api/matches/{id}/state`,
-`/api/matches/{id}/timeline`, replay, catch-up, and assistant routes. Demo
-fixtures remain available when `VITE_DEMO_MODE` is not `false`.
-
-## Offline experiments and deployment
-
-`app/fine_tuning/train.py` contains the PEFT/LoRA training entry point and
-`evaluate.py` reports measured accuracy, macro-F1, and a confusion matrix.
-Training is intentionally offline and the API does not train at startup.
-`app/rag/ingestion.py` prepares idempotent document chunks for Qdrant.
-
-Build the Cloud Run image with `docker build -t matchlens-api backend` and pass
-secrets as environment variables. Netlify should use the same
-`VITE_API_BASE_URL` setting and an SPA fallback for `index.html`. Render can
-run the same Docker image with the configured `PORT`.
diff --git a/backend/recent.zip b/recent.zip
similarity index 100%
rename from backend/recent.zip
rename to recent.zip
diff --git a/backend/requirements.txt b/requirements.txt
similarity index 100%
rename from backend/requirements.txt
rename to requirements.txt
diff --git a/backend/scripts/check_qdrant.py b/scripts/check_qdrant.py
similarity index 100%
rename from backend/scripts/check_qdrant.py
rename to scripts/check_qdrant.py
diff --git a/backend/scripts/download_cricsheet.py b/scripts/download_cricsheet.py
similarity index 100%
rename from backend/scripts/download_cricsheet.py
rename to scripts/download_cricsheet.py
diff --git a/backend/scripts/ingest_knowledge.py b/scripts/ingest_knowledge.py
similarity index 100%
rename from backend/scripts/ingest_knowledge.py
rename to scripts/ingest_knowledge.py
diff --git a/backend/tests/test_api.py b/tests/test_api.py
similarity index 100%
rename from backend/tests/test_api.py
rename to tests/test_api.py
diff --git a/backend/tests/test_cricket.py b/tests/test_cricket.py
similarity index 100%
rename from backend/tests/test_cricket.py
rename to tests/test_cricket.py
diff --git a/backend/tests/test_partnership.py b/tests/test_partnership.py
similarity index 100%
rename from backend/tests/test_partnership.py
rename to tests/test_partnership.py